A New Algorithm For Assessing The Xco2 Over Peninsular Malaysia Based On Gosat Data
Loading...
Date
2015-10
Authors
CHONG KEAT, SIM
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The increasing carbon dioxide (CO2) concentration induced by anthropogenic activities has been the focal point of many studies due to the adverse effects of global warming and climate change on the environment. To achieve a healthy environment, studying the transport, distributions and source regions of CO2 in Malaysia is necessary. The main purpose of this research is to develop an algorithm for calculating the column-averaged dry air mole fraction of carbon dioxide (XCO2) over Peninsular Malaysia. Four regression algorithms, which are denoted as XCO2 NEM, XCO2 SWM, PCA1 (XCO2 NEM season) and PCA2 (XCO2 SWM season), were developed using Greenhouse Gases Observing Satellite (GOSAT) data and statistical methods. In addition, this study seeks to analyse and investigate the impacts of selected atmospheric variables with the XCO2 data. Different statistical analysis methods, including multiple linear regression (MLR) and principal component regression (PCR), were applied to the GOSAT datasets. Additional analysis was conducted in different monsoon seasons to achieve this study’s objective. SPSS software was used to test the performance of the MLR and PCR methods in terms of the root-mean-square-error (RMSE). The results showed that the XCO2 regression equations using the MLR method were highly correlated with atmospheric variables in the NEM (R= 0.826, R2 = 0.682) and SWM (R= 0.802, R2 = 0.643) seasons. The validation results showed that XCO2 yielded a strong R2 for the NEM and SWM seasons, i.e., 0.8035 to 0.8156 and 0.8093 to 0.8178, respectively. Additionally, for the PCR method, the best fit results for the XCO2 data gave the
high adjusted R2 coefficients, i.e., 0.898 dan 0.868 for both the NEM and SWM seasons. The common variables that appeared in both the PCA1 and PCA2 equations were the AOT and temperature. The obtained validation results exhibited high coefficients of determination for the NEM and SWM seasons, i.e., 0.8584 to 0.9149 and 0.8832 to 0.8944, respectively. The RMSE for the predicted XCO2 values using the MLR method were 1.56208 and 1.71421 for the NEM and SWM, respectively, and the corresponding RMSEs were 0.84924 and 1.01879, respectively with PCR method. The predicted and observed XCO2 values exhibited very good agreement in term of consistency and reliability of the prediction model. The PCR method resulted in better predicted XCO2 values over peninsular Malaysia than the MLR method. Overall, these results clearly indicate the advantage of using GOSAT data and a correlation analysis to investigate the impact of atmospheric variables on XCO2 over peninsular Malaysia. Therefore, this modelling approach has great potential in other areas.
Description
Keywords
A New Algorithm For Assessing The Xco2 , Over Peninsular Malaysia Based On Gosat Data